conductivity in contrast to nanopowders and the photoluminescence spectra was further employed to examine phase purity.
nitrogen and the formation temperature also indicates the minimum thermal requirements reported in current CuAlO2 literatures.
The multi-step thermodynamic formation and multiphase during solid state reaction may give rise to phase purity issues and long dwelling time. In addition, the low oxygen partial pressure may also increase the processing complexity and affect the thermal stability. This may pose less challenges for thin film fabrication. However, for bulk CuAlO2 synthesis, it is necessary to ease the complexities during solid state reaction. The effect of applied electric field during ceramic sintering has been studied in many aspects, ranging from sintering morphologies, phase formation to sintering kinetics. Most of the thermodynamic studies on CuAlO2 and other delafossites often use oxygen partial pressure as a crucial variable to determine the critical transformation temperatures. The applied electric field and related electric effects may modify the kinetics and should therefore be investigated when considering bulk CuAlO2 fabrication. In addition, any tunable properties enabled by electric field may also shed new light on the current p-type TCO study.
Figure 50. Thermogravimetric diagrams of heating CuO-Al2O3 powders in both air (a) and nitrogen (b), with Arrhenius plot of rate constant k as a function of
temperature based on reaction (4).
Spark plasma sintering (SPS, FCT Systeme GmbH) was used to provide high DC current and low voltage under a low vacuum condition (7×10-2 torr). The low vacuum is believed to lower the synthesis temperature for CuAlO2 according to previous TG/DTA tests. Even though the SPS is inherently favorable for fabricating CuAlO2, there have been few reports on synthesizing or sintering CuAlO2 by SPS or under an electric field.
In order to better examine the effect of electric field, a homemade field-assisted annealing (FAA) furnace was used in which two Pt electrodes were inserted into a tube furnace with two ends applied on the two sides of the cold-pressed pellets. In this case, there are no pressure and vacuuming so the sintering could be performed in ambient atmosphere. A
low voltage of 50V and high voltage of 2kV were applied respectively to generate different DC electric fields. During the course of studying reactive synthesis under electric field, the sinterability and phase stability of pure CuAlO2 were also investigated, as shown in the XRD patterns from Figure 51(a). It is noteworthy that even though the as-synthesized CuAlO2 is thermally stable upon heating cycles to high temperatures, it becomes unstable under the electric field, in both SPS and FAA scenarios. The CuAlO2
powders used in this study were prepared by a sol-gel method107. The redox of Cu is severe when sintered by SPS and a total decomposition into Cu and Al2O3 occurred at a temperature of 1050 ºC as shown in Figure 51(a). The powder remained unsintered at temperatures below 800 ºC. Therefore, the consolidation of CuAlO2 by SPS is not feasible even at low vacuum levels. For the field-assisted annealing (FAA) and sintering in ambient atmosphere without pressure, decomposition was also observed with resultants the same as that from the reverse reaction of (20). Even though the oxygen partial pressure in this case is high, making the reverse reaction likely to happen, the CuAlO2 and Cu(I) are more stable at high temperatures above 1000 ºC. One of the major difference between SPS and FAA is the plasma discharge within the particles at low temperatures prior to sintering and this may lead to Cu redox and melting at grain interfaces, which could explain the phase configurations of CuAlO2 sample sintered by SPS. The FAA on the other hand, could also exhibit large electric field with space charge and higher activation energies at interfaces and grain boundaries. Previous studies168,169 on electric-field assisted sintering also observed the high surface energies and local melting at intermediate electric field, especially for smaller particles. As a consequence, the presence of CuO might also partially result from the oxidation of Cu after elimination of the interfaces and gaps during coarsening and sintering.
Figure 51. XRD patters after sintering by SPS and FAA, with two different starting powders: (a) As-synthesized CuAlO2 and (b) CuO-Al2O3 mixed powders.
Given the difficulties of consolidating CuAlO2 particles, CuO-Al2O3 mixed powders could be an alternative providing oxide powder beds to avoid redox and shortened the overall preparation procedures. As shown in Figure 51(b), pure delafossite CuAlO2 could be synthesized and sintered simultaneously through SPS at 800 ºC for only 10min under 80MPA. In comparison to the TG/DTA curves in flowing nitrogen with lower oxygen partial pressure, the onset CuAlO2 forming temperature in SPS is even lower. Even in the case of FAA samples processed in air, the CuAlO2 could be formed at a temperature 50 ºC lower than that indicated in the TG/DTA curves. There are also some CuO and Al2O3
yet not spinel CuAl2O4 presented. This could be due to the secondary decomposition of as-formed CuAlO2 due to the long dwelling time required to consolidate the pellets in the pressureless sintering.
Figure 52. SEM images of samples sintered from CuO-Al2O3 precursors. SPS: (a) 700 ºC, 10min, 80MPa; (b) 800 ºC, 10min, 80MPa; (c) 1000 ºC, 10min, 80MPa;
FAA: (d) no electric field applied, 1000 ºC, 10hr; (e) 10kV/cm,1000 ºC, 10hr; (f) 0.25kV/cm, 1000 ºC, 10hr.
The corresponding SE micrographs exhibit evident variances between the two sintering techniques. Figure 52(a)-(c) illustrate the densification of reactive synthesized CuAlO2 particles at elevated sintering temperatures. Laminar grains were observed in Fig. 3c in a direction perpendicular to the uniaxial press. The inset of Figure 52(c) shows the onset of lamination at the surface. The surface layer is Cu-rich from the EDS spectra, indicating a partial Cu redox might occur near the surface when SPS sintering at higher temperatures. The outside layer of the SPS sample in contact with graphite mold and vacuum surroundings may be more likely to undergo the Cu redox reaction than the inner parts. The release of oxygen from the solid-state reaction during sintering could ease the low oxygen partial pressure in the inner parts, thus facilitating the formation of delafossite and prevent the Cu reduction. The FAA treatment in both low and high electric field yielded different microstructures. When high electric field was applied, the lath-shaped domain structure was formed whereas a fish-scale-shaped microstructure was
formed at low electric field. The white particles residing in the grain boundaries is alumina rich. Those microstructures with aligned or partially aligned grains may also be beneficial to its anisotropic electrical properties.
The as-sintered samples all exhibit semiconducting behavior depicted by Figure 53(a).
Due to the different phase configurations the conductivity varies, with the Cu-rich sample showing the highest conductivity, followed by the phase-pure CuAlO2 by SPS reactive sintering at 800 ºC. The FAA treated samples show lower conductivity because of the presence of alumina phase. Additionally, with the phase separation and the increase of heterointerfaces, the DC conductivity was impaired significantly. Table VII also summarizes some electrical properties obtained by Van der Pauw–Hall method. The conductivity measured from four-point probe is consistent with the observation from Figure 53(a). All the samples exhibit p-type conductivity and the carrier densities decrease significantly as the increase portion of non-conducting alumina.
Table VII. Electrical Properties of Selected Samples Measured by Hall Measurement
Sintering method - Starting
powder σ (Scm-1) RH (cm3C-1) np (cm-3)
CuAlO2 powder pellet 0.029 +8.65 7.22×1017
SPS-CuAlO2 0.427 +5.24 1.19×1018
SPS-CuO-Al2O3-800 ºC 0.132 +4.38 1.43×1018
SPS-CuO-Al2O3-1000 ºC 0.220 +3.24 1.93×1018
FAA-10kV/cm- CuO-Al2O3 0.005 +12.8 4.88×1017
FAA-0.25kV/cm- CuO-Al2O3 0.007 +14.5 4.3×1017
The CuAlO2 could also emit light in the UV-blue regions at room temperatures due to the near-band-edge excitation, which could be regarded as a sensitive approach to confirm the phase purity and examine the crystallinity. The photoluminescence emission spectra shown in Figure 53(b) summarized the samples with different phase configurations involved in this study. When comparing the phase pure bulk CuAlO2 with sol-gel synthesized nanopowders, the emission peak from the bulk sample is well defined with a slight red-shift due to the weak quantum confinement119. There is a small should
peak at around 380nm in the nanopowders, which could be ascribed to the intrinsic luminescent band from alumina. In the bulk samples, only the broad and intensified CuAlO2 direct transition was presented. There are no other emission bands in the visible region identified in both samples. In Figure 53(c), the pressureless sintered sample in high electric field shows two emission bands at 460nm and 485nm respectively.
Considering the bandgap of CuO (1.6eV), Cu2O(2.3eV) and CuAl2O4 (2.1eV)170, it is unlikely for those compounds to emit in this region. Therefore, the observed shortened Stokes shift in this sample after long dwelling in air might be originated from Cu2+ ions embedded in CuAlO2 or alumina matrix. There is a tiny band at ~600nm which could be attributed to the near-band-edge emission from CuAl2O4 phase. This phase amount may not be sufficient to be revealed by the XRD, however the PL spectra could be utilized to identify if any metastable spinel phase remains after heating over 1000 ºC. Figure 53(d) corresponds to the CuAlO2 powder sintered by SPS. The dominant peak centered at
~540nm correspond to the existence of Cu2O, which is not observed in FAA and reactive SPS samples. This may lead to the similar conjecture that the direct SPS sintering of CuAlO2 could expedite the reduction of Cu(II) and the metastable Cu(I) after high temperature sintering may reside in either CuxO or alumina after the total decomposition of the CuAlO2.
Figure 53. (a) Temperature dependent conductivity (sintering technique-starting powder) and (b) room-temperature photoluminescence emission spectra.